JP2014519062A - Storage system and logical unit management method thereof - Google Patents

Abstract

To minimize the size of a management information page for storing format management information and to reduce the management size of the management information page.
A controller for managing a management information logical unit area for storing format management information for managing whether or not an access target logical unit has been formatted is managed in a front page area and a rear page area. When formatting the logical unit for management, the capacity of the format management information is calculated based on the capacity of the logical unit for access, and the number of pages of management information is calculated from the ratio of the calculated format management information capacity and the page management unit. And the format management information of one page is assigned to the front page area, the storage destination is managed by the front page area address, the format management information of two or more pages is assigned to the rear page area, and the storage destination is assigned to the rear page area. Manage by area address and page length.
[Selection] Figure 6

Description

  The present invention relates to a storage system and a logical unit management method for managing logical units formed on storage areas of a plurality of storage devices.

  2. Description of the Related Art In recent years, in storage systems that handle large-scale data, functions of the storage system itself have been expanded in order to store a large amount of data that increases every day in a large number of hard disk drives (HDDs). For this reason. Increasing the number of logical volumes formed on physical storage areas of a large number of hard disk drives is also being carried out. As the number of logical volumes increases, management information for managing logical volumes also needs to be increased. The management information of the logical volume is usually stored in a semiconductor memory, but the semiconductor memory has a small capacity and is expensive.

  Therefore, logical volume management information is stored in RAID (Redundant Array of Inexpensive Disks) composed of low-cost hard disk drives, and the expansion of the storage system scale is achieved at a low price (Patent Literature). 1).

  On the other hand, in the case of a storage system having a plurality of storage control devices, the hard disk drive is accessed from each storage control device, and thus access exclusion is required. As a method related to exclusive control between controllers, for example, a method has been proposed in which a part of a cache memory and a logical volume are allocated to each controller to eliminate exclusive control between the controllers with respect to the cache memory (see Patent Document 2). ).

JP 2008-0665706 A JP 2006-1114064 A

  Patent Document 2 employs a method of managing a storage area of a cache memory by dividing it into segments. This method is also used for managing a storage area of a storage device such as a hard disk drive (HDD) that constitutes a RAID. However, when the storage area of a storage device such as a hard disk drive (HDD) constituting a RAID is divided and managed by a specific management unit (page), the following problems arise.

  For example, when storing logical volume management information in units of pages, no matter how small the management information is, the management information must always be allocated to one page as a storage area for storing management information. For this reason, if small management information is stored in a page-by-page storage area when the page size is large, a useless storage area is generated. Conversely, when the management information is large and the page size is small, the number of pages for storing the management information increases.

  The present invention has been made in view of the above-described problems of the prior art, and its purpose is to minimize the size of the management information page for storing the format management information and reduce the management size of the management information page. It is an object of the present invention to provide a storage system and a logical unit management method thereof.

  To achieve the above object, the present invention is a logical unit formed in a storage area of a plurality of storage devices, wherein one or a plurality of access target logical units to be accessed by an access request source have been formatted. A management information logical unit area that stores format management information for managing whether or not the management information logical unit area is divided into a front page area and a rear page area, and the controller includes the access target When executing the format processing for the logical unit for use, the capacity of the format management information is calculated based on the capacity of the logical unit for access, and the calculated capacity of the format management information and the management unit of the page of the specific capacity Ratio management information required for managing the format management information in the management unit. A portion of the format management information corresponding to the calculated number of pages of the management information page is allocated to the front page area, and the page of the calculated management information page Allocating the remaining format management information among the format management information corresponding to the number to the rear page area, and storing the execution result of the format process as the format management information in the front page area or the rear page area. Features. At this time, the controller manages the storage location of the format management information allocated to the front page area by a front page area address, and stores the storage location of the format management information allocated to the rear page area as a rear page area. Can be managed by address and page length.

  According to the present invention, it is possible to minimize the size of the management information page for storing the format management information and reduce the management size of the management information page.

It is a block diagram for demonstrating the relationship between normal LU and a difference bitmap table. FIG. 6 is a configuration diagram of a management information LU. It is a block diagram for demonstrating the process in which access competes between controllers. FIG. 3 is a configuration diagram of a management information LU to which a stripe column is assigned for each management unit. It is a block diagram which shows the whole structure of a computer system. It is an internal block diagram of a disk array device. It is a block diagram of a local memory. It is a block diagram of a drive management table. It is a block diagram of an RG management table. It is a block diagram of an LU management table. It is a block diagram of a QF management table. It is a block diagram of the virtual memory management table for management information. It is a block diagram of the management information page management table. It is a flowchart for demonstrating the process of the controller which received the quick format instruction | indication. It is a flowchart for demonstrating the process at the time of ensuring a difference bitmap table. It is a flowchart for demonstrating the virtual memory allocation process for management information. It is a flowchart for demonstrating a front page area | region allocation process. It is a flowchart for demonstrating a back page area | region allocation process. It is a figure for demonstrating the process for allocating a back page area | region to LU arrange | positioned on a RAID group. It is a flowchart for demonstrating the QF registration process performed with a polling opportunity. It is a flowchart for demonstrating a job starting process. It is a flowchart for demonstrating a difference bitmap update process. It is a flowchart for demonstrating a next format area | region search process. It is a flowchart for demonstrating the virtual memory access process for management information. It is a flowchart for demonstrating the process which calculates a physical address from a virtual memory address. It is a flowchart for demonstrating a 1st charge core determination process. It is a flowchart for demonstrating a 2nd charge core determination process. It is a flowchart for demonstrating a read command process. It is a flowchart for demonstrating normal read processing. It is a flowchart for demonstrating a cross call read process. It is a block diagram of 2nd Example, Comprising: It is a block diagram in the case of preserve | saving a difference bitmap table in a thin provisioning pool.

(Summary of Invention)
For example, when a user volume is constructed on a RAID configured with a plurality of storage devices, for example, a controller performs a format process, for example, a quick format process, on the storage area of each storage device.

  At this time, as shown in FIG. 1, among the normal LUs 10 which are LUs to be subjected to the quick format process and are configured on the RAID, the format target area 12 is divided into a plurality of quick formats in a 1 MB quick format management unit. It is divided into target areas 14, and whether or not each quick format target area 14 has been formatted is determined based on the difference bitmap table 16 (hereinafter, a format for managing whether or not the normal LU 10 indicating the logical unit for access has been formatted). A table in which management information is registered with difference bits is referred to as a difference bitmap table).

  In this case, if the quick format target area 14 has been formatted, “0” is stored in the difference bitmap table 16 corresponding to the quick format target area 14, and the quick format target area 14 is in an unformatted state. In this case, “1” is stored in the difference bitmap table 16 corresponding to the quick format target area 14.

  That is, whether or not each quick format target area 14 has been formatted is managed by the difference bitmap table 16 using a 1-bit difference bit.

  When the size of the normal LU 10 is 7680 TB, the capacity of the differential bitmap table 16 is about 1 GB.

  In order to store 1 GB of data in a semiconductor memory, a large amount of semiconductor memory is required.

  Specifically, when formatting a storage device, a difference bitmap table in which management information for managing the presence / absence of formatting for each logical unit (hereinafter sometimes referred to as LU (Logical Unit)) is registered. The capacity is 1 bit per 1 MB (megabyte) of LU. Therefore, when the total capacity of all LUs subject to quick formatting is 7680 TB (terabytes), the capacity of the difference bitmap table is about 1 GB (gigabytes) for all LUs.

  Here, when managing the presence / absence of the format processing of each LU using the differential bitmap table with 1 GB / (page management unit) = number of pages for management information, the smaller the page management unit, the more The number of information pages will increase.

  For example, the simplest management method is considered on the assumption that there is a maximum capacity of 128 TB per LU and that there are a maximum of 4096 LUs. When the management method of managing all management information pages allocated per LU is adopted, the size of the difference bitmap table is as follows, assuming that the management information (address information) of one management information page is 4 bytes. It becomes.

  First, the size of the difference bitmap table for a maximum of 128 TB is 16 MB. That is, when 128TB is divided by 1MB of the bitmap management unit, and the value obtained by the division is managed by 1 bit per 1MB, and converted to bytes and managed, the size of the difference bitmap table for 128TB Will be 16MB.

  When the management unit is small and the page management unit is 4 MB, the number of management information pages that can store the 16 MB differential bitmap table is 4 management information pages.

  Since the management information of one management information page is 4 bytes, the amount of information necessary for managing the four management information pages is 16 bytes.

  The memory size required for managing the information amount of 16 bytes for 4096 LUs is 64 KB.

  Next, when the page management unit is 1 MB, the number of management information pages that can store a 16 MB differential bitmap table is 16 management information pages.

  The amount of information necessary for managing the 16 management information page is 16 × 4 bytes = 64 bytes.

  The memory size necessary to manage the amount of information of 64 bytes for 4096 LUs is 256 KB.

  That is, the smaller the management unit is, the larger the number of management information pages is, and accordingly, the size of the memory for storing the management information is increased.

  On the other hand, when the page management unit is large, for example, when the page management unit is 10 MB and the capacity of the LU to be subjected to quick formatting is 128 TB, the management information necessary for storing the 16 MB differential bitmap table The number of pages for use is 16 MB / 10 MB = 1.6. If the fraction is rounded up, the page becomes 2 management information.

  That is, when the size of the difference bitmap table is 16 MB, the number of management information pages necessary for storing this capacity is 1.6 pages in calculation. However, since the management information page is managed for each management information page, the number of pages necessary for storing the difference bitmap table is two management information pages. Therefore, a storage area that is not used is generated in the last management information page of the two management information pages. As described above, the larger the page management unit, the more useless storage areas that are not actually used are generated.

  Therefore, as an LU for storing the data of the difference bitmap table 16, as shown in FIG. 2, a management information LU 18 is formed on a RAID group composed of a plurality of storage devices, and the management information LU 18 is stored in the management information LU 18. The data of the difference bitmap table 16 is stored.

  At this time, the area of the management information LU 18 is divided into a front page area 20 and a rear page area 22.

  When the data of the differential bitmap table 16 is stored in the management information LU 18, the data of the differential bitmap table 16 is information for each LU. Depending on the LU configuration, the differential bitmap corresponding to each LU is stored. The size of the table 16 changes. For this reason, in order to save the data of the difference bitmap table 16 in the management information LU 18, it is necessary to divide and manage for each LU.

  Therefore, the capacity of the difference bitmap table 16 is managed by using a management information page obtained by dividing the capacity of the difference bitmap table 16 by a specific capacity, for example, 1 MB page management unit. At this time, the management information page is set to an integral multiple of the stripe column size, which is the minimum unit of read access or write access.

  In the management information LU 18, an area calculated by the capacity of the maximum number of user volumes in the RAID × page size (size of management information page) is allocated to the front page area 20.

  Further, an area required as management information calculated from the size of the entire RAID group is allocated to the rear page area 22.

  Thereafter, in the process of dividing the normal LU 10 into a plurality of LUs, the number of pages required to store the data of the differential bitmap table 16 is calculated for each LU based on the capacity of the divided LUs. . When the calculated number of pages is 2 or more, the data of the difference bitmap table 16 corresponding to one page is assigned to the front page area 20, and the data of the difference bitmap table 16 after the second page is assigned to the rear page area 22. .

  At this time, the storage destination of the data allocated to the front page area 20 is managed by the front page area address, and the storage destination of the data allocated to the rear page area 22 is allocated to the rear page area address and the rear page area 22. Managed by the page length of the specified page.

  Further, as shown in FIG. 3, when there are data 26 and 28 of different pages in the plurality of stripe columns 24 formed in the management information LU 18, one controller is determined for each page. When two controllers (storage control devices) access the same stripe row 24, an access conflict occurs.

  Therefore, as shown in FIG. 4, an integer multiple of the stripe column 24 constructed in the management information LU 18 is defined as the management information page 30, and the stripe column 24 constructed in the management information LU 18 is defined for each management information page 30. In addition to managing them collectively, a responsible controller is set for each management information page 30. For example, in the management information page 30, ownership information indicating that the access to the management information page 30 is occupied is stored in association with one of the controllers. Thereby, it is possible to prevent an access conflict between the controllers.

(First embodiment)
This embodiment has a controller that manages the area of the management information logical unit that stores the format management information for managing whether the access target logical unit is formatted in the front page area and the rear page area. When formatting the logical unit for access, the capacity of the format management information is calculated based on the capacity of the logical unit for access, and the ratio of the calculated format management information capacity to the page management unit The number of pages is calculated, the format management information of one page is assigned to the front page area, the storage destination is managed by the front page area address, the format management information of two or more pages is assigned to the rear page area, and the storage destination is Management is based on the backward page area address and page length.

  A first embodiment of the present invention will be described below with reference to the drawings.

  FIG. 5 is a block diagram showing the overall configuration of the computer system. 5, the computer system includes a management server 50, a disk array device 52, a network 54, and a plurality of host computers (hereinafter sometimes referred to as hosts) 56.

  Each host 56 is a computer device provided with information processing resources such as a CPU (Central Processing Unit) and a memory, and includes, for example, a personal computer, a workstation, a mainframe, and the like. Each host 56 issues an access request designating an LU or logical volume provided from the disk array device (storage subsystem) 52, for example, a write request or a read request to the disk array device (storage subsystem) 52. The LU or logical volume can be accessed.

  Each host 56 includes an information input device (not shown) such as a keyboard, a switch, a pointing device, and a microphone, and an information output device (not shown) such as a monitor display and a speaker.

  The network 54 includes, for example, a SAN (Storage Area Network), a LAN (Local Area Network), the Internet, a public line, or a dedicated line. For example, when the network 54 is a SAN, communication between each host 56 and the disk array device 52 via the network 54 is performed according to the fiber channel protocol, and when the network 54 is a LAN, This is performed according to the TCP / IP (Transmission Control Protocol / Internet Protocol) protocol.

  The disk array device 52 is configured as a storage system having a storage control device 58 and a plurality of additional enclosures (expansion housings) 60.

  Next, FIG. 6 shows an internal configuration diagram of the disk array device.

  6, the disk array device 52 includes a storage control device 58 having a plurality of controllers 70 and 72 and a basic housing 74, and a plurality of storage devices 76 are housed in the basic housing 74. A plurality of storage devices 76 are also stored in each additional chassis 60.

  Examples of the storage device 76 include storage devices such as a hard disk device, a semiconductor memory device, an optical disk device, a magneto-optical disk device, a magnetic tape device, and a flexible disk device, and these storage devices are devices that can read and write data. It is.

  When using hard disk devices as storage devices, for example, FC (Fibre Channel) disks, SCSI (Small Computer System Interface) disks, SATA (Serial ATA) disks, ATA (AT Attachment) disks, SAS (Serial Attached SCSI) disks, etc. Can be used.

  When using a semiconductor memory device as a storage device, SSD (Solid State Drive) (flash memory), FeRAM (Ferroelectric Random Access Memory), MRAM (Magnetoresistive Random Access Memory), phase change memory (Ovonic Unified Memory), RRAM (registered trademark) ) (Resistance Random Access Memory) or the like.

  Further, each storage device 76 can form a RAID group, for example, RAID 4, RAID 5, RAID 6, etc., or each storage device 76 can be divided into a plurality of RAID groups. At this time, a plurality of logical units (LU) and a plurality of logical volumes can be formed on the physical storage area of each storage device 76.

  The LU is a logical unit provided to the host 56 as an access target of the host 56, and is divided into a normal LU and a virtual LU.

  The normal LU is composed of logical storage areas formed on the storage device 76. On the other hand, the virtual LU is provided by the thin provisioning function, and is composed of storage unit units called pages. In the stage before the data at the time of creating the virtual LU is written, a logical storage area formed from the physical storage area is not associated with the page. Thereafter, when new data is written to the page, a partial storage area of the logical storage area formed on the storage device 76 is allocated to the page on which the writing has been performed, and this allocated storage is performed. Data is stored in the area.

  A LUN (Logical Unit Number) is assigned as an identifier to the normal LU and the virtual LU, and a logical block address LBA (Logical Block Address) is assigned to each segment. At this time, each host 56 sends a command including a logical address made up of an identifier LUN and a logical block address LBA to the controller 70 or the controller 72 of the disk array device 52, whereby a storage area corresponding to a normal LU or a virtual LU. Can access the data stored in

  The controller 70 includes an MPU 80, a memory controller (MC) 82, a local memory (LM) 84, a cache memory (CM) 86, a host interface 88, a data transfer control circuit (DCTL) 90, a disk interface 92, It is configured as a # 0 controller having an expander (EXP) 94.

  The MPU 80 is configured by a microprocessor that performs overall control of the entire controller 70, and this microprocessor is configured by, for example, multiple cores.

  The memory controller 82 performs a mapping process on the local memory 84 based on a command from the MPU 80.

  The local memory 84 is configured as a storage device that stores system configuration information, management information, various control programs, and the like.

  The cache memory 86 is configured as a storage device that temporarily stores user data input and output to and from the disk array device 52.

  The host interface 88 includes a port (not shown) for connecting the disk array device 52 to the network 54, interprets various commands transmitted from each host 56, executes processing in accordance with the commands, and And exchange data.

  The data transfer control circuit 90 is connected to the memory controller 82, cache memory 86, host interface 88, disk interface 92, and data transfer control circuit 110, and controls data transfer to these connected parts.

  The disk interface 92 exchanges data with the storage device 76 or the expander 94 housed in the expansion chassis 76 and functions as an interface for performing protocol control during communication.

  At this time, the disk interface 92 reads data from the storage device 76 in the basic chassis 74 or reads data from the storage device 76 in the expansion chassis 60 via the expander 94 based on read access from the MPU 80. Each read data is transferred to the data transfer control circuit 90. Furthermore, the disk interface 92 stores write data in the storage device 76 in the basic chassis 74 or the storage device 76 in the expansion chassis 60 via the expander 94 based on the write access from the MPU 80.

  The expander 94 is connected to the disk interface 92 and is also connected to the storage device 76 in the basic chassis 74 and the storage device 60 in the extension chassis 76, and data from the disk interface 92 is transferred to the basic chassis 74 or the extension chassis 60. It is comprised with the branch circuit which branches.

  The controller 72 includes an MPU 100, a memory controller (MC) 102, a local memory (LM) 104, a cache memory (CM) 106, a host interface 108, a data transfer control circuit (DCTL) 110, and a disk interface 112. , And an expander (EXP) 114.

  The MPU 100 is composed of a microprocessor that performs overall control of the entire controller 72, and this microprocessor is composed of, for example, multiple cores.

  The memory controller (MC) 82 performs mapping processing for the local memory 104 based on a command from the MPU 100.

  The local memory 104 is configured as a storage device that stores system configuration information, management information, various control programs, and the like.

  The cache memory 106 is configured as a storage device that temporarily stores user data input and output to and from the disk array device 52.

  The host interface 108 includes a port (not shown) for connecting the disk array device 52 to the network 54, interprets various commands transmitted from each host 56, executes processing according to the command, and And exchange data.

  The data transfer control circuit 110 is connected to the memory controller 102, the cache memory 106, the host interface 108, the disk interface 112, and the data transfer control circuit 90, and controls data transfer to these connected parts.

  The disk interface 112 functions as an interface for exchanging data with the storage device 76 or the expander 114 housed in the expansion chassis 60 and performing protocol control during communication.

  At this time, the disk interface 112 reads data from the storage device 76 in the basic chassis 74 or reads data from the storage device 76 in the expansion chassis 60 via the expander 114 based on read access from the MPU 100. Each read data is transferred to the data transfer control circuit 110. Further, the disk interface 112 stores the write data in the storage device 76 in the basic chassis 74 or the storage device 76 in the expansion chassis 60 via the expander 114 based on the write access from the MPU 110.

  The expander 114 is connected to the disk interface 112 and is connected to the storage device 76 in the basic chassis 74 and the storage device 76 in the expansion chassis 60, and the data from the disk interface 112 is transferred to the basic chassis 74 or the expansion chassis 60. It is comprised with the branch circuit which branches.

  Next, FIG. 7 shows a configuration diagram of the local memory.

  In FIG. 7, local memories 84 and 104 include a drive management table 120, an RG (RAID group) management table 122, an LU management table 124, a QF (quick format) management table 126, and a management information virtual memory management. A table 128 and a management information page management table 130 are stored.

  Next, FIG. 8 shows a configuration diagram of the drive management table.

  In FIG. 8, the drive management table 20 is a table for managing the storage device 76 housed in the basic chassis 74 or the expansion chassis 76, and includes a plurality of HDU (Hard Disk Unit) entries 140. Each HDU entry 140 stores information about an additional chassis number (extended enclosure number) 142, an HDU number 144, a capacity 146, a type 148, and a status 150.

  For example, “0” is stored in the additional chassis number (extended chassis number) 142 as the number of the additional chassis 76. For example, “0” is stored in the HDU number 144 as the HDU number. For example, “300 GB” is stored in the capacity 146 as the capacity of the HDU. For example, “SAS” is stored in the type 148 as the type of the storage device 76.

  In the state 150, “Normal” is stored when the storage device 76 is normal.

  Next, FIG. 9 shows a configuration diagram of the RG management table.

  In FIG. 9, the RG management table 122 is a table for managing the storage devices 76 constituting the RAID group, and includes a plurality of RG entries 160. Each entry 160 stores information about the RG number 162, RAID level 164, type 166, capacity 168, free capacity 170, and status 172.

  For example, “00” is stored in the RG number 162 as the RAID group number, and “RAID” 5 (4D + 1P) is stored in the RAID level 164.

  The type 166 stores “SAS” as the type of the storage device 76. In the capacity 168, “1.2 TB” is stored as the capacity of the RAID group. In the free capacity 170, “900 GB” is stored as an unused capacity. In the state 172, “Normal” is stored when it is normal and formatted.

  Next, FIG. 10 shows a configuration diagram of the LU management table 124.

  In FIG. 10, an LU management table 124 is a table for managing a target LU that is a target of quick formatting, and includes a plurality of LU entries 180.

  Each LU entry 180 stores information regarding the LU number 182, the belonging RG number 184, the capacity 186, the stripe size 188, the status 190, and the ownership 192.

  For example, “0000” is stored in the LU number 182 as a number for identifying the target LU that is the target of the quick format. In the affiliation RG number 184, “00” is stored as the RG number to which the target LU belongs. In the capacity 186, “200 GB” is stored as the capacity of the target LU. In the stripe size 188, “64 KB” is stored as the size of the stripe column assigned to the target LU.

  In the state 190, “Normal” is stored when the target LU is normal. The ownership 192 stores information related to the controller or core (core constituting the MPU 80 or MPU 100) that occupies the access right to the target LU. For example, “owner 192” stores “control device # 1” in the case of the controller 72, and “control device # 0” in the case of the controller 70.

  Next, FIG. 11 shows a configuration diagram of the QF management table 126.

  In FIG. 11, a QF management table 126 is a table for managing the status of quick formatting and the progress status of quick formatting, and includes a plurality of LU entries 200. Each LU entry 200 stores information about the LU number 202 of the target LU to be subjected to quick formatting, the QF progress rate 204 of the target LU, and the virtual memory number 206.

  For example, “0000” is stored in the LU number 202 as the number of the target LU to be subjected to quick formatting. In the QF progress rate 204, “100%” is stored as the quick format progress rate of the target LU that is the target of the quick format when the quick format is completed.

  For example, “0000” is stored in the virtual memory number 206 as a number for accessing the management information virtual memory management table 128.

  Next, FIG. 12 shows a configuration diagram of the management information virtual memory management table 128.

  In FIG. 12, a management information virtual memory management table 128 is a table for managing the differential bitmap table 16 stored in the management information LU 18 and is composed of a plurality of virtual memory entries 210. Each virtual memory entry 210 stores information regarding the target LU 212, the storage destination LU 214, the front page area address 216, the rear page area address 218, and the page length 220.

  For example, “LU0000” is stored in the target LU 212 as information for specifying the target LU that is the target of the quick format.

  In the storage destination LU 214, “LU 4096” is stored as information for specifying the management information LU 18 in which the differential bitmap table 16 is stored.

  For example, “0x00001000” is stored in the front page area address 216 as an address assigned to the front page area 20.

  For example, “0x00011000” is stored in the rear page area address 218 as an address assigned to the rear page area 22.

  In the page length 220, “3” is stored as information on the length of the page allocated to the rear page area 22, for example, when the page length allocated to the rear page area 22 is three pages. Is done.

  Next, FIG. 13 shows a configuration diagram of the management information page management table 130.

  In FIG. 13, the management information page management table 130 is a table for managing the area of the management information LU 18 in units of management information pages 30 (pages for management information), and from a plurality of LU 4096 addresses 230. Composed. As the LU 4096 address 230, “0x00000000” to “0x00FFF000” are assigned.

  Each LU 4096 address 230 stores information on the storage destination LU 232, the storage destination address 234, the allocated virtual memory 236, and the ownership 238.

  For example, “4096” is stored in the storage destination LU 232 as the storage destination LU number for storing the differential bitmap table 16.

  For example, “0x00000000” is stored in the storage destination address 234 as an address corresponding to the LU 4096 address 230.

  For example, “0000” is stored in the allocated virtual memory 236 as information for specifying the virtual memory entry 210 of the management information virtual memory management table 128.

  The ownership 238 stores information related to the controller or core (core constituting the MPU 80 or MPU 100) that occupies the storage destination LU. For example, “owner 238” stores “control device # 1” in the case of the controller 72, and “control device # 0” in the case of the controller 70.

  Next, the quick format process will be described with reference to the flowchart of FIG.

  This process is started when a quick format process is instructed from the management server 50 to the controller 70 or the controller 72.

  For example, when the management server 50 instructs the controller 70 to perform a quick format process, the MPU 80 of the controller 70 issues a request for prohibiting execution of a process such as a configuration change as an exclusive process between the controllers. The data is transferred to the controller 72 via 90 (S11).

  Thereafter, the MPU 80 executes a process of updating information in the QF management table 126 (S12).

  For example, the MPU 80 executes processing for initializing the QF progress rate 204 of the QF management table 126 and the like.

  Next, the MPU 80 executes a process for securing an area for storing the difference bitmap table 116 (S13).

  Next, the MPU 80 executes a process for registering the quick format process result in the cache memories 86 and 106 (S14).

  Thereafter, the MPU 80 executes a process for ending the exclusion process between controls (S15), returns the process result to the management server 50, and ends the process in this routine.

  Next, processing for securing an area for storing the difference bitmap table will be described with reference to the flowchart of FIG.

  This process is a process performed in step S13 of FIG.

  The MPU 80 executes processing for allocating the management information virtual memory (S21), and ends the processing in this routine.

  Next, the management information virtual memory allocation process will be described with reference to the flowchart of FIG.

  This process is a process performed in step S21 of FIG.

  The MPU 80 calculates the required capacity from the LU capacity (S31).

  For example, when the target LU to be subjected to the quick format is a normal LU 10 and the capacity of the normal LU 10 is 128 TB, the MPU 80 sets the capacity (required capacity) necessary for storing the differential bitmap table 16. , Calculated as 16MB.

  Next, the MPU 80 calculates the required number of pages from the required capacity (S32). For example, if the required capacity is 16 MB and the page management unit is 1 MB, the required capacity / page management unit = number of pages is calculated, and the required page number = 16 management information pages.

  Next, the MPU 80 performs allocation to the front page area 20 (S33). For example, when the number of pages is 16 management information pages, a process for allocating data of the difference bitmap table 16 corresponding to one management information page to the front page area 20 is performed.

  Next, the MPU 80 determines whether the required number of pages is 2 or more (S34). If it is determined in step S34 that the required number of pages is one management information page, the MPU 80 responds to the management server 50 that the required number of pages is one management information page, and the processing in this routine is performed. Exit.

  On the other hand, if it is determined in step S34 that the required number of pages is equal to or larger than the two management information pages, the MPU 80 allocates the data of the difference bitmap table 16 after the second management information page to the rear page area 22. (S35), the processing result is returned to the management server 50, and the processing in this routine is terminated.

  Next, the front page area allocation processing will be described with reference to the flowchart of FIG.

  This process is a process performed in step S33 of FIG.

  The MPU 80 refers to the front page area 20 of the management information LU 18, searches for an unused management information page (free area) in the front page area 20 (S41), and searches for the unused management information page thus searched. Are allocated as an area for storing data of the difference bitmap table 16 (S42).

  Next, the MPU 80 writes the data of one management information page among the data of the difference bitmap table 16 to the allocated unused management information page, and the address of the management information page to which the data is written. Is registered in the management information virtual memory management table 128 as the front page area address 216 (S43), the processing result is returned to the management server 50, and the processing in this routine is terminated.

  At this time, the MPU 80 can also register a use flag indicating assigned in the virtual memory entry 210 of the management information virtual memory management table 128.

  Next, rear page area allocation processing will be described with reference to the flowchart of FIG.

  This process is a process performed in step S35 of FIG.

  The MPU 80 refers to the arrangement of the LU on the RAID group and provisionally allocates the rear page area 22 to which the target LU that is the target of the quick format belongs (S51).

  For example, as shown in FIG. 19, when the 0000th LU, the 0001th LU, and the 0002th LU are arranged as LUs on the RAID group, the target LUs to be subjected to the quick format. Is the 0001th LU, the page 22A and a part of the page 22B are provisionally assigned to the 0002th LU as the pages temporarily assigned to the rear page area 22, and the page 22B is assigned to the 0001th LU. , A part of page 22C and a part of page 22D are temporarily allocated, and a part of page 22D is temporarily allocated to the 0000th LU.

  Thereafter, the MPU 80 calculates the number of pages of the management information page 30 necessary for storing the data of the differential bitmap table 16 of each LU based on the capacity of each LU. As the number of pages, for example, when the LU of 0000 is 1 page, the LU of 0001 is 3 pages, and the LU of 0002 is 2 pages, the first page of each LU is the front page area 20. Already allocated to page 0, page 0 is allocated to the 0000th LU, page 2 is allocated to the 0001 LU, and page 1 is allocated to the 0002th LU. Assign.

  In other words, the MPU 80 assigns each temporarily assigned LU to the rear page area 22 in a left-justified manner based on the calculated number of pages (S52).

  In this case, as shown in FIG. 19, the MPU 80 assigns the page 22A to the 0002th LU as the management information page 30 to be assigned to the rear page area 22, and the pages 22B and 22C to the 0001th LU. Assign. Note that the page 22D is an unused area.

  Thereafter, the MPU 80 registers the addresses of the pages 22A, 22B, and 22C allocated to the rear page area 22 as the management information page 30 of each LU in the management information virtual memory management table 128 as the rear page area address 218. At the same time, the page length 220 of each page 22A, 22B, 22C is registered (S53), the processing result is returned to the management server 50, and the processing in this routine is terminated.

  Next, specific processing contents of the quick format registration will be described with reference to the flowchart of FIG.

  This process is a process performed in step S14 of FIG.

  First, polling is performed between the controller 70 and the controller 72, and exclusive processing for performing quick format registration is executed by one of the controllers 70 and 72 (S61).

  Here, when the quick format registration is performed by the controller 70, the MPU 80 determines whether there is an LU registered in the quick format (QF), that is, an LU to be processed in the quick format (S62). .

  If it is determined in step S62 that there is no LU to be processed by the quick format, the MPU 80 proceeds to the process of step S65. On the other hand, if it is determined in step S62 that there is an LU subject to the quick format process, the MPU 80 refers to the LU management table 124 and determines whether or not the controller in charge of the quick format process is the controller 70. Is determined (S63).

  When the MPU 80 determines in step S63 that the other controller, that is, the controller 72 is the controller responsible for the quick format process, the MPU 80 proceeds to the process in step S65. In step S63, the controller 70 determines that the quick format process is performed. If it is determined that the controller is in charge of the process, the process for starting the quick format job is executed (S64).

  When the MPU 80 is configured with multiple cores, any one of the multiple cores executes the processes of Steps S61 to S63, and in Step S63, the cores that execute the processes of Steps S61 to S63. However, it is determined whether or not the core in charge of the quick format processing is another core or the own core (the core that has executed the processing in steps S61 to S63).

  Thereafter, the MPU 80 completes the exclusion process between the controllers (S65) and ends the process in this routine.

  Next, job activation processing will be described with reference to the flowchart of FIG.

  This process is a process performed in step S64 of FIG.

  The MPU 80 executes initialization setting processing for initializing variables and the like stored in the local memory 84 (S71), and thereafter executes extent lock securing processing (S72).

  For example, the MPU 80 executes an exclusive process for executing only the quick format process as a job.

  Next, the MPU 80 executes a one-difference format process (S73).

  For example, when the normal LU 10 is the target of the quick format process, the MPU 80 converts the registered data of one quick format target area 14 from “1” to “0” as the format process for one quick format target area 14. To change to "".

  Next, the MPU 80 determines whether or not the one-difference formatting process has been completed (S74). If it is determined in step S74 that the one-difference formatting process has not been completed, the process proceeds to step S73. Returning, if it is determined that the one-difference formatting process is complete, a differential bitmap update process is performed (S75).

  At this time, the MPU 80 also performs a write process on the management information virtual memory in the course of performing the differential bitmap update process.

  Next, the MPU 80 executes the process for releasing the extent lock (S76), performs the next format area search process (S77), and determines that the next format area is present in step S77, sets the next format area. (S78), and then the process returns to step S72 to repeat the processes of steps S72 to S77.

  On the other hand, if it is determined in step S77 that the formatting process for all areas has been completed, the MPU 80 completes the formatting process for the normal LU 10 (S79), and ends the process in this routine.

  Next, the contents of the differential bitmap update process will be described with reference to the flowchart of FIG.

  This process is a process performed in step S75 of FIG.

  The MPU 80 instructs to update the management information virtual memory. For example, when the quick format processing for the quick format target area 14 of the normal LU 10 is completed, the MPU 80 stores the difference bitmap table 16 corresponding to the quick format target area 14. Processing for changing the difference bit from “1” to “0” is performed (S81), and the processing in this routine is terminated.

  Next, the next format area search process will be described with reference to the flowchart of FIG.

  This process is a process performed in step S77 of FIG.

  The MPU 80 refers to the difference bitmap table 16 as a management information virtual memory reference instruction process, and performs a process for determining a difference bit (S91).

  Next, the MPU 80 determines whether or not formatting is performed in the next format area (S92).

  At this time, if the MPU 80 refers to the difference bitmap table 16 and determines that the next format area has been formatted, the MPU 80 returns to the process of step S91 and determines that the next format area is unformatted. If it is determined that all the format areas have been formatted, the processing in this routine is terminated.

  Next, the management information virtual memory access processing will be described with reference to the flowchart of FIG.

  This process is performed in step S81 in FIG. 22 or step S91 in FIG.

  The MPU 80 performs processing for specifying the difference bitmap table 16 based on the surface number (the LU that is the target of the quick format and specifying the virtual memory entry 210) and calculating the physical address from the virtual memory address. Implement (S102).

  Next, the MPU 80 refers to the cache memory 86 based on the calculated physical address, and determines whether or not the data registered in the differential bitmap table 16 exists on the cache memory 86 (S103).

  If the MPU 80 determines in step S103 that the data registered in the differential bitmap table 16 exists in the cache memory 86, the MPU 80 proceeds to the process of step S105, and is registered in the differential bitmap table 16 in step S103. If it is determined that the data does not exist on the cache memory 86, the data registered in the differential bitmap table 16 is read from the storage device 76 onto the cache memory 86 (S104).

  Thereafter, the MPU 80 performs reference processing or update processing on the data registered in the difference bitmap table 16 on the cache memory 86 (S105), returns the processing result to the management server 50, and performs processing in this routine. Exit.

  Next, processing for calculating a physical address from a virtual memory address will be described with reference to the flowchart of FIG.

  This process is a process performed in step S102 of FIG.

  The MPU 80 determines, based on the management information virtual memory address obtained from the command, the management information virtual memory address / page management unit ( 1MB) is calculated, and based on the number of pages of management information page 30 obtained by this calculation, it is determined whether or not it is on the first management information page (S111).

  If it is determined in step S111 that the management information page 30 is the first page, the MPU 80 calculates the physical address of the management information page 30 by offset calculation based on the front page area address 216 (S112).

  On the other hand, if it is determined in step S111 that the management information page 30 is the second page or later, the MPU 80 physically calculates the management information page 30 for the second and subsequent pages by offset calculation based on the rear page area address 218. An address is calculated (S113).

  After the process of step S112 or step S113, the MPU 80 returns each process result to the management server 50, and ends the process in this routine.

  Next, processing when determining ownership of the management information page will be described.

  Normally, the ownership for accessing the management information page 30 is owned by the controller that manages the LU to which the management information page 30 is assigned or the core belonging to the controller. For this reason, the core or controller that accesses the management information page 30 is limited to the core or controller that accesses the management information page 30. Therefore, in the normal access processing to the management information virtual memory, The ownership of the page 30 is not determined.

  On the other hand, when the recovery process of the storage device 76 is performed, the ownership of the management information page 30 is determined. There are the following two methods for determining the ownership of the management information page 30.

  Next, the first responsible core determination process for determining the ownership of the management information page will be described with reference to the flowchart of FIG.

  For example, the MPU 80 of the controller 70 refers to the management information virtual memory management table 128 on the basis of the recovery instruction from the management server 50, and determines the virtual memory address from the physical address assigned to the management information page 30 to be recovered. (S121), and the target LU 212 to be restored is calculated based on the calculated virtual memory address (S122).

  Next, the MPU 80 refers to the LU management table 124 based on the calculated target LU 212, identifies the responsible core from the ownership 192 of the LU management table 124 (S123), returns the processing result to the management server 50, The processing in this routine is terminated.

  Next, the second assigned core determination process for determining the ownership of the management information page will be described with reference to the flowchart of FIG.

  First, when the controller 70 is instructed for recovery processing from the management server 50, the MPU 80 of the controller 70 refers to the management information page management table 130 in order to search for the LU to be recovered, and becomes the recovery target. If the LU exists as the storage destination LU 232, the responsible core is determined based on the ownership 238 registered in the LU 4096 address 230 (S131), the processing result is returned to the management server 50, and the processing in this routine is terminated. To do.

  In this case, since the responsible core can be directly identified simply by referring to the management information page management table 130, the process for determining the ownership of the management information page rather than the first responsible core determination process. Time is shortened.

  Next, read processing for a normal LU will be described with reference to the flowchart of FIG.

  When the controller 70 receives a read command from the host 56, the MPU 80 of the controller 70 refers to the ownership 192 of the LU management table 124 based on the read command and determines whether the controller is in charge of processing by the read command. (S141).

  If it is determined in step S141 that the controller should be in charge of read command processing, the MPU 80 executes normal read processing (S141), and ends the processing in this routine.

  On the other hand, if it is determined in step S141 that the controller that is to handle the read command is another controller, the MPU 80 transfers the read command to the controller 72 via the data transfer control circuit 90. The process is executed (S143), and the process in this routine is terminated.

  Next, normal read processing will be described with reference to the flowchart of FIG.

  This process is a process performed in step S142 of FIG.

  First, the MPU 80 determines whether or not the read destination LU specified by the read command among the plurality of LUs existing in the normal LU 10 has been formatted (S151).

  At this time, the MPU 80 refers to the difference bitmap table 16 corresponding to the LU of the read destination, determines whether or not the quick format target area 14 of the read destination has been formatted, and if it is unformatted, The format activation process is started (S152), and the process proceeds to step S153. If it is determined in step S151 that the format has been completed, the process proceeds to step S153.

  In step S153, the MPU 80 executes extent lock as an exclusive process for executing read processing only on the read destination LU, and thereafter, a segment for securing a read destination area on the cache memory 86. The securing process is executed (S154).

  Next, the MPU 80 determines whether it is a cache hit or a cache miss (S155).

  In other words, the MPU 80 determines whether or not the read data exists on the cache memory 86. If it is determined that there is a cache miss, the MPU 80 performs staging processing for reading the read data from the storage device 76 onto the cache memory 86. (S156), and the process proceeds to step S157.

  On the other hand, if the MPU 80 determines in step S155 that it is a cache hit, the MPU 80 executes a data transfer process for transferring the read data existing on the cache memory 86 or the read data staged from the storage device to the host 56 ( S157).

  Thereafter, the MPU 80 performs a process for releasing the segment (S158), subsequently executes a process for releasing the extent lock (S159), and ends the process in this routine.

  Next, the cross call read process will be described with reference to the flowchart of FIG.

  This process is a process performed in step S143 of FIG.

  When receiving the read command transferred from the controller 70, the MPU 100 of the controller 72 determines whether or not the read destination LU has been formatted based on the received read command (S161).

  At this time, the MPU 100 refers to the difference bitmap table 16 corresponding to the LU of the read destination, determines whether or not the quick format target area 14 of the read destination has been formatted, and if it is unformatted, The format activation process is started (S162), and the process proceeds to step S163. If it is determined in step S161 that the format has been completed, the process proceeds to step S163.

  In step S163, the MPU 100 executes extent lock as an exclusive process for executing read processing only on the read destination LU, and thereafter, a segment for securing a read destination area on the cache memory 106. The securing process is executed (S164).

  Next, the MPU 100 determines whether it is a cache hit or a cache miss (S165).

  That is, the MPU 100 determines whether or not the read data exists on the cache memory 106. If the MPU 100 determines that there is a cache miss, the MPU 100 transfers the read data from the storage device 76 to the cache memory 106 and the cache memory 84. The double writing staging process to be read respectively is executed (S166), and the process proceeds to step S168.

  On the other hand, if the MPU 100 determines in step S165 that there is a cache hit, the MPU 100 executes the inter-control cache copy process of copying the read data existing on the cache memory 106 to the cache memory 84 of the controller 70, and proceeds to the process of step S168. To do.

  Thereafter, the MPU 100 executes data transfer processing for transferring the read data existing on the cache memory 106 or the read data staged from the storage device 76 to the host 56 (S168).

  Next, the MPU 100 performs a process for releasing a segment (S169), subsequently executes a process for releasing an extent lock (S170), and ends the process in this routine.

  As described above, the determination as to whether or not the formatting has been completed is executed on the controller side in charge of processing of the LU to be subjected to the quick formatting. In addition, since the process of referring to the information of the difference bitmap table 16 is also executed by the controller in charge of the quick format process, the management information page 30 is in charge of the normal quick format process or read access from the host 56. There is no need to determine the controller or core, and access is exclusive between controllers or cores.

  Further, although the case where the controller 70 receives a read command from the host 56 has been described, when the controller 70 or the controller 72 receives a write command from the host 56, the normal write process is executed instead of the normal read process. In addition, by executing the cross call write process instead of the cross call read process, a process for writing the write data to the write destination LU can be executed.

  At this time, if the controller 70 receives a read command or write command from the access request source host 56, the virtual memory address indicating the address on the differential bitmap table (format management information) 16 based on the read command or write command. And the number of pages of the management information page 30 corresponding to the access destination is calculated from the ratio between the calculated virtual memory address and the page management unit (1 MB), and the management information for one page is calculated as the calculated number of pages. When the page 30 exists, the physical address of the access destination is calculated from the front page area address corresponding to the management information page 30 of one page, and is specified by the front page area address corresponding to the management information page 30 of one page. The difference bitmap table 16 stored in the forward page area 20 On condition that the formatted information already is registered, it executes the read access or write access to the calculated physical address accessed.

  Further, when the controller 70 calculates the number of pages of the management information page 30 corresponding to the access destination, if there are two or more management information pages 30 in the calculated number of pages, the controller 70 manages two or more pages. The physical address of the access destination is calculated from the rear page area address and page length corresponding to the information page 30, and the rear specified by the rear page area address and page length corresponding to the management information page 30 of two or more pages On the condition that the formatted information is registered in the differential bitmap table 16 stored in the page area 22, the read access or the write access is executed with respect to the access destination of the calculated physical address.

  In the start of the format process in step S152 and the format start process in step S162, the job start process in FIG. 21 is performed.

  According to the present embodiment, it is possible to minimize the size of the management information page 30 that stores the difference bitmap table (format management information) 16 and to reduce the management size of the management information page 30.

  According to the present embodiment, the format management information (difference bitmap table 16) for managing the management information LU 18 includes the front page area address 216, the rear page area address 218, regardless of the page management unit. Management can be performed using information on the page length 220 of the management information page 30 allocated to the rear page area 22. Therefore, the format management information can be managed in the minimum stripe column size unit as the access unit of the RAID group, and a useless area generated for exclusive control can be minimized.

  Further, since the format management information for managing the management information LU 18 is constant regardless of the size of the LU to be formatted, it is possible to cope with further capacity expansion of the disk array device.

  (Second embodiment)

  FIG. 31 shows a configuration diagram when the storage destination of the difference bitmap table 16 is a thin provisioning pool as a second embodiment of the present invention.

  In FIG. 31, when a thin provisioning pool 500 is formed on logical storage areas of a plurality of storage devices 76, a plurality of chunks 502 are formed in the thin provisioning pool 500, and a virtual volume 504 is assigned to any one of the chunks 502. Can do.

  When the virtual volume 504 is not accessed from the host 56, no logical volume or page is allocated to the virtual storage area. On the other hand, for example, when a write access is input as an access request from the host 56 to the virtual volume 504, a capacity virtualization control function (a processing function by starting a capacity virtualization control program) is performed at the write access timing. The virtual storage area of each virtual volume 504 is divided into a plurality of blocks, and a logical volume registered in the thin provisioning pool 500 is assigned to each block.

  At this time, when the virtual volume 504 is allocated to any one of the chunks 502, the page 506 of the virtual volume 504 is composed of, for example, 32 MB. It cannot be used for.

  On the other hand, when the page 510 to be allocated to the chunk 502 is divided into a plurality of chunks 502 in units of stripe column size with the management information page size = stripe column size and the thin provisioning pool 500 is divided into a plurality of chunks 502, a plurality of chunks 502 have a plurality of Pages 510 can be allocated without creating free space. Thereby, the thin provisioning pool 500 can be shared with the user's virtual LU. As the capacity of the chunk 502, for example, 1 GB can be used.

  According to this embodiment, by dividing the thin provisioning pool 500 into a plurality of chunks 502 in units of stripe column size, a plurality of pages 510 can be allocated in each chunk 502 without forming a free area. .

  Note that the management information LU 18 that is the storage destination of the differential bitmap table 16 can also be formed in an SSD constituting a RAID. In this case, the access speed can be increased more than when the management information LU 18 is formed in the HDD.

  The present invention is not limited to the above-described embodiments, and includes various modifications. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Further, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

  Further, each of the above-described configurations, functions, and the like may be realized by hardware by designing a part or all of them with, for example, an integrated circuit. Each of the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by the processor. Information such as programs, tables, and files for realizing each function is stored in a memory, a hard disk, a recording device such as an SSD (Solid State Drive), an IC (Integrated Circuit) card, an SD (Secure Digital) memory card, a DVD ( Digital Versatile Disc) can be recorded and placed.

  10 Normal LU, 14 Quick Format Target Area, 16 Differential Bitmap Table, 18 Management Information LU, 20 Front Page Area, 22 Back Page Area, 24 Stripe Column, 30 Management Unit, 50 Management Server, 52 Disk Array Device, 54 Network, 56 hosts, 58 storage controllers, 60 additional chassis, 70, 72 controllers, 74 basic chassis, 76 storage devices, 80, 100 MPU, 84, 104 local memory, 86, 106 cache memory, 88, 108 host interface, 90, 110 Data transfer control circuit, 92, 112 Disk interface, 120 Drive management table, 122 RG management table, 124 LU management table, 126 QF management table, 128 Management information virtual memory management table, 1 Page management table for the 0 management information.

Claims (12)

A plurality of storage devices, logical units formed in storage areas of the plurality of storage devices, one or a plurality of access target logical units to be accessed by an access request source, and storage of the plurality of storage devices Management information, which is a logical unit formed in an area, in which an area for storing format management information for managing whether or not the access-target logical unit has been formatted is divided into a front page area and a rear page area A logical unit, and one or a plurality of controllers for controlling the data input / output for each logical unit and executing the formatting process for the logical unit for access,
The controller is
When the format process is performed on the access target logical unit, the format management information capacity is calculated based on the access target logical unit capacity, and the calculated format management information capacity and page management of the specific capacity are managed. The number of management information pages required to manage the format management information in the management unit of the page is calculated from the ratio to the unit, and the format management corresponding to the calculated page number of the management information page A part of the format management information is allocated to the front page area, the remaining format management information of the format management information corresponding to the calculated number of pages of management information is allocated to the rear page area, The format management result is used as the format management information. The storage system to be stored in the front page region or the rear page area. The storage system according to claim 1,
The controller is
When there is one management information page in the calculated number of pages, the storage location of the format management information corresponding to the one management information page is allocated to the front page area, and the one management information page The storage location of the format management information corresponding to is managed by a front page area address, and when there are two or more management information pages in the calculated number of pages, the format corresponding to the two or more management information pages A management information storage destination is allocated to the rear page area, and a storage destination of the format management information corresponding to the two or more management information pages is set as a rear page area address and a page length of the two or more management information pages. A storage system that is managed by The storage system according to claim 1,
The management information page allocated to the front page area or the rear page area is a minimum unit of read access or write access by the controller, and is an integer of stripe columns formed in the logical storage area of the storage device A storage system that is configured with double the capacity. The storage system according to claim 1,
The storage system is characterized in that the management information page stores, in association with any one of the controllers, ownership information indicating that the access to the management information page is occupied. The storage system according to claim 1,
The controller is
When a read command or a write command is received from the access request source, an access destination is specified based on the read command or the write command, and the access destination in the format management information stored in the management information logical unit is specified. Referencing the corresponding format management information, and executing read access or write access to the access destination on condition that the formatted information is registered in the format management information corresponding to the access destination And storage system. The storage system according to claim 2,
The controller is
When a read command or a write command is received from the access request source, a virtual memory address indicating an address on the format management information is calculated based on the read command or the write command, and the calculated virtual memory address and the page The number of pages of management information corresponding to the access destination is calculated from the ratio with the management unit, and when one management information page exists in the calculated number of pages, it corresponds to the one management information page. The physical address of the access destination is calculated from the front page area address, and formatted into the format management information stored in the front page area specified by the front page area address corresponding to the one management information page On the condition that the above information is registered. When a read access or a write access is executed for an access destination with no access, and there are two or more management information pages in the calculated number of pages, the rear page area address corresponding to the two or more management information pages The format management stored in the rear page area specified by the page length and the rear page area address corresponding to the two or more management information pages is calculated from the page length. A storage system, wherein a read access or a write access is executed with respect to the access destination of the calculated physical address on the condition that formatted information is registered in the information. A plurality of storage devices, logical units formed in storage areas of the plurality of storage devices, one or a plurality of access target logical units to be accessed by an access request source, and storage of the plurality of storage devices Management information, which is a logical unit formed in an area, in which an area for storing format management information for managing whether or not the access-target logical unit has been formatted is divided into a front page area and a rear page area A logical unit management method for a storage system, comprising: a logical unit for controlling data input / output to each logical unit, and one or a plurality of controllers for executing format processing for the logical unit for access ,
As a step when the controller executes the format processing for the logical unit for access,
The controller calculating the capacity of the format management information based on the capacity of the logical unit for access;
Based on the ratio between the calculated format management information capacity and the specific page management unit, the controller calculates the number of management information pages required to manage the format management information in the page management unit. A calculating step;
The controller assigning a portion of the format management information of the format management information corresponding to the calculated number of pages of management information to the front page area;
The controller assigning the remaining format management information among the format management information corresponding to the calculated number of pages of management information to the rear page area;
A logical unit management method for a storage system, comprising: a step of storing the execution result of the formatting process in the front page area or the rear page area as the format management information. A storage unit logical unit management method according to claim 7,
When there is one management information page for the calculated number of pages, the controller allocates a storage location of the format management information corresponding to the one management information page to the front page area, and Managing the storage location of the format management information corresponding to the management information page by the front page area address;
When there are two or more management information pages in the calculated number of pages, the controller assigns the storage location of the format management information corresponding to the two or more management information pages to the rear page area, and A logical unit management method for a storage system, wherein a storage destination of the format management information corresponding to two or more management information pages is managed by a rear page area address and a page length of the two or more management information pages . A storage unit logical unit management method according to claim 7,
The management information page allocated to the front page area or the rear page area is a minimum unit of read access or write access by the controller, and is an integer of stripe columns formed in the logical storage area of the storage device A logical unit management method for a storage system, characterized in that it is configured with a double capacity. A storage unit logical unit management method according to claim 7,
Logical unit management of a storage system, wherein the management information page stores, in association with any one of the controllers, ownership information indicating that the access to the management information page is occupied Method. A storage unit logical unit management method according to claim 7,
As a step of processing when the controller receives a read command or a write command from the access request source,
The controller specifying an access destination based on the read command or the write command;
The controller refers to the format management information corresponding to the access destination among the format management information stored in the management information logical unit;
The controller executing a read access or a write access to the access destination on condition that formatted information is registered in the format management information corresponding to the access destination. A storage unit logical unit management method. A storage system logical unit management method according to claim 8,
As a step of processing when the controller receives a read command or a write command from the access request source,
The controller calculates a virtual memory address indicating an address on the format management information based on the read command or the write command;
The controller calculates a page number of management information pages corresponding to an access destination from a ratio between the calculated virtual memory address and the management unit of the page;
The controller calculating the physical address of the access destination from the front page area address corresponding to the one management information page when there is one management information page in the calculated number of pages;
On condition that the controller has registered formatted information in the format management information stored in the front page area specified by the front page area address corresponding to the one management information page. Executing read access or write access to the access destination of the calculated physical address;
When there are two or more management information pages in the calculated number of pages, the controller determines the physical address of the access destination from the rear page area address and the page length corresponding to the two or more management information pages. A calculating step;
The controller registers formatted information in the format management information stored in the backward page area specified by the backward page area address and the page length corresponding to the two or more management information pages. And a step of executing read access or write access to the access destination of the calculated physical address on the condition of the above.

Images